Cerebral palsy is the most common motor deficit in children with an incidence of 2-3 per 1000 live births. An important socioeconomic problem is that 44% of children with cerebral palsy (CP) never learn to walk independently even with current treatments. Non-ambulatory children with CP have higher risk of hip dysplasia, scoliosis, osteoporosis, diminished growth and respiratory illness. Ironically, hope for better functional outcome is reduced for these children because they are commonly excluded from motor control research and clinical innovations. For example, current locomotor interventions (body weight supported treadmill training (BWSTT), assistive devices, robotic locomotor training) are designed for higher functioning children who have attained upright control of the trunk. Thus locomotor intervention is either inaccessible or inefficient for those children with the greatest need. Researchers working with young typically developing (TD) infants (who have poor trunk control, naturally), tip the babies' trunks forward when supported on a treadmill to help them step. A specific angle of inclination has not, however, been proposed nor has the rationale for this posture been articulated or tested. It is hypothesized that this technique reduces the demands for trunk control while simultaneously reducing the muscle force required for infants to swing their legs and step. It is proposed that this technique could also work for children with moderate-to-severe CP. Using a custom trunk support device with adjustable forward inclination angle and pediatric motorized treadmill, step responses will be documented monthly in 7 infants (AIM 1) during the age range when trunk postural control normally develops (1-7 mo) and in 30 children with moderate-to-severe CP (Gross Motor Function Classification System Level III, IV and V) (2-4 or 6-8 years of age) (AIM 2). Concurrent assessment of the segmental level of trunk control will provide information necessary to test the hypotheses that the optimal angle of trunk inclination is related to the level of trunk control available to the infant/child. This is the first study to systematically evaluate the influence of trunk control on stepping responses during BWSTT. Surface electromyography (EMG) of trunk and leg muscles will be recorded during studies for AIM 1 and 2, to test the hypothesis that providing trunk support will reduce the postural demands and enhance reciprocal activation patterns for leg and trunk muscles during treadmill-elicited stepping (AIM 3). Insights from this research will provide specific information to optimize performance during locomotor treatment for those children with CP who have the greatest need.